cambridgezoologyfandomcom-20200215-history
What evidence is there if any for homeothermy and endothermy in cynodonts and early mammals? Suggest why they might have arisen.
Most modern mammals are endothermic homeotherms, maintaining a constant body temperature by producing heat via metabolic activity. Reptiles, including early synapsids such as pelycosaurs are ectothermic, and though they may be homeotherms, they have no internal source of body heat. It is difficult to assess the temperature physiology of fossil animals as this physiology is based in the soft tissues. However, osteological and ecological correlates can be used ranging from nasal passage structure to predator:prey ratio. These indicate that the earliest mammals, like those today, were endothermic. This also appears to have been the case with the closest cynodont relatives of the mammals, but the endothermic status of less mammal-like cynodonts, and less yet non-cynodont therapsids is more debated, though a gradual transition may be evidenced. Homeothermy is important in order to allow proper functioning of metabolic processes and maintenance of reasonable biological complexity. However, the reasons for the evolution of endothermy are less clear. Various hypotheses have been put forward, divisible into those based on constant temperature as the key driver, and those with the higher aerobic activity endothermy allows being the main selective advantage. Such a divided approach may well not be necessary however, with various selective pressures possibly leading ultimately to evolution of mammalian endothermy. Homeothermy is the maintenance of a constant body temperature. All animals must do this to some extent as the complex biochemical reactions of metazoans, and more so vertebrates, rely on relatively constant internal conditions. In modern terrestrial vertebrates this is achieved by two key means – behavioural homeothermy, where animals move to areas of different ambient temperatures in order to maintain a roughly constant body temperature, and metabolic homeothermy, where maintenance of a constant body temperature is achieved by metabolic heat production either from basal metabolic rate or from muscular action. The opposite situation from homeothermy is poikilothermy, where an animal tolerates a large range of different temperatures, though this division is far from clear cut, and even poikilothermic animals must maintain a degree of thermal constancy. Endothermy is the situation where body heat is generated my internal means, not originating from the external environment – its opposite is ectothermy. Metabolically based homeothermy is thus inherently endothermic, but even animals which are poikilothermic may also be endotherms – for example hummingbirds, the temperature of which greatly decreases at night, though the body heat used to maintain activity during the day is metabolically (i.e. internally) generated. These animals massively reduce their metabolic rate when resting – they are bradymetabolic – whereas most mammals and birds do not, and are tachymetabolic. Being extremely small, and thus having a very large surface area:volume ratio, hummingbirds would lose too much heat were they to maintain high metabolism at night. Except when hibernating, tachymetabolism is the norm in all but the tiniest mammals, as it incurs less risk if it can be supported. Considering that modern reptiles generally display behavioural homeothermy, it is a fair assumption that the reptilian ancestor of modern mammals did also. There is direct fossil evidence for ectothermic homeothermy in pelycosaurs, with Dimetrodon and Edaphosaurus both possessing large sails on their back spanning between extended neural arches, which were almost certainly used for thermoregulation (Fig.X). They would have speeded up the warming up process after dawn, and also allowed loss of heat were the animal in danger of overheating. The blood flow to these sails could likely be regulated to modify heat transfer levels. Modern mammals are almost all endothermic homeotherms, and even the earliest fossil mammals such as Morganucodon appear to have most of the osteological features associated with endothermy. The situation in the therapsids is somewhat less clear. Later cynodonts appear to have also had many of the osteological correlates of endothermy, but in animals such as Thrinaxodon whether these are yet present is a matter of some debate. The main means used by palaeontologists to asses whether fossil synapsids possessed endothermy have been hard tissue correlates, as this is mostly all that is preserved. Hair is a key mammalian trait and is not found in any other groups. It – like feathers in birds - is very much indicative of endothermy as it is necessary in order to retain the heat produced by metabolism and thereby prevent massive metabolic wastage. Unfortunately however, hair does not fossilize at all well as it is constructed from protein, and there are no stem mammal fossils with hair, though it is present in fossils of some animals within the Mammalia such as Eomaia scansoria (Ji et al., 2002). Thus it cannot be used to accurately assess the timing of initial development of endothermy. Key osteological features used to data endothermy have often been those associated with higher aerobic capacity. Endothermic animals have on average a 10 times higher aerobic capacity than do ectotherms (Kemp, 2005). This means that a lizard, for example, though it can run as fast as a mammal, has much lower stamina and must recuperate for some time after running in order to pay back its oxygen debt via lactic acid metabolism. Higher aerobic capacity requires and allows several morphological features. The secondary palate in mammals is in part thought to assist in allowing breathing whilst consuming food (Fig.X), thereby allowing air intake to be maintained in order to provide enough oxygen to maintain high aerobic capacity. Modern mammals also have a vertebral column clearly differentiated into lumbar and thoracic sections, with only the latter having ribs. These ribs are in turn fused to form a rib cage which can be moved by the diaphragm to allow greatly increased breathing efficiency, and thereby the oxygen supply needed to maintain aerobic capacity (Fig.X). When these adaptations appeared is not clear cut, with most of the vertebral column still possessing ribs in even early mammals, though a diaphragm likely was present in most if not all cynodonts, and certainly in the earliest mammals (Kemp, 2005). The development of the secondary palate is also gradual, with cynodonts such as Thrinaxodon displaying a partially closed secondary palate (Fig.X), but with mammal-like cynodonts such as thritheledonts and all early mammals themselves possessing a full secondary palate. The movement of the legs under the body away from the sprawling gate of pelycosaurs and other reptiles is also likely a key correlate to aerobic metabolism, facilitating breathing using the diaphragm. It will have avoided Carrier’s constraint, which is an interference to breathing caused by the legs when running in lizards (Fig.X). In lizards breathing whilst running is more difficult than in animals with their legs below their body, as the lungs are compressed alternately by leg movement. However, the importance of this issue is likely overstated as lizards are thought in fact to indeed be able to breathe whilst running to an extent and to increase their breathing rate 3-4 times above resting (Kemp, 2005). Despite this, legs under the body are certainly an indicator of a highly active lifestyle, probably one relying on high aerobic capacity. Even the most basal therapsids possess the mammalian ankle joint (Fig.X) and other adaptations to upright walking such as movement of muscular attachment sites more dorsally on the limb girdles. However, the mammals and their close cynodont relatives have taken this morphotype to an extreme, becoming highly skeletally agile. Maxilloturbinals in the nose (Fig.X) are thought also to be correlated with increased aerobic activity as the increased breathing rate which such activity demands would otherwise lead water loss being too extensive. These turbinals allow water in exhaled air to recondense, thus preventing dehydration. These are present in Morganucodon, but, despite much being made of the evidence, only small transverse ridges which could possible correlate to them, or cartilaginous versions thereof, are visible in cynodonts (Kemp, 2006). Thus from this evidence alone, one can only securely hypothesize homeothermy to have occurred in the mammals themselves, but it is possible that these ridges may represent a sort of ‘proto-turbinal’ phase and this could in turn correlate to a gradual acquisition of endothermy. Increased aerobic activity as well as the increase in basal metabolic rate (BMR) associated with endothermy need large amounts of fuel. Thus adaptations which improve feeding efficiency can also be regarded as possibly correlating with endothermy. Increasingly differentiated dentition is seen in the therapsids through time, with a variety of very specialized forms arising. For example, gorgonopsids have long piercing canines for puncturing and holding food combined with interdigitating incisors for cutting (Fig.X). Such interdigitation is absent in basal pelycosaurs for example, but is yet more elaborate in carnivorous cynodonts such as Thrinaxodon. Equally sophisticated herbivorous dentition is seen, with the traversodontid cynodonts such as Exaeretodon having broad flat teeth with sharp cutting surfaces for shearing up plant material. In early mammals such as Morganucodon the teeth are highly sophisticated, differentiated, and precisely occluding (Fig.X). These tooth patterns indicate a necessity to improve feeding efficiency in order to fuel a more metabolically active lifestyle. By the time of the latest cynodonts such as the tritheledontids (likely the sister group of mammals), the dentition is as sophisticated as in early mammals – though the lack of a strong dentary-squamosal jaw hinge likely prevented great pressure from being applied whilst feeding. If one thus goes by dental morphology, what is seen is a gradual trend towards higher metabolism, reaching its peak in the cynodonts immediately ancestral to the mammals. A similar trend is seen in the muscular system driving jaw action. In pelycosaurs one observes the origination of the temporal fenestra, allowing greater muscle attachment and thereby increased force to be applied whilst feeding (Fig.X). This is taken much further by the time of the cynodonts, with many having an extremely large temporal fenestra, so much so that its ventral/lateral side is reduced to a small zygomatic arch and the skull to a sagittal crest (Fig.X). This allows not only much greater forces, but also accuracy and precision of jaw movement, which correlates to the precisely occluding teeth. Brain size is also thought to correlate positively with endothermic homeothermy, with one of the characteristics of mammals being a much enlarged brain compared to reptiles. This is difficult to assess from fossils as in earlier synapsids the brain does not fill the cavity, but can at least be estimated. Although theoretically via behaviour one is able to maintain a reasonably constant temperature, the level of constancy possible with endothermy is much greater. This allows evolution of greater complexity, as complex biological systems tend to be more temperature sensitive, hence bacteria being able to thrive at 90°C (Pörtner, 2004). The central nervous system (CNS) is highly sensitive to temperature and is the first structure to fail on temperature reduction. Thus in order to have the large brain they do, maintenance of body temperature is essential for modern mammals. For what purpose this large mammalian brain originally evolved is something of an uncertainty, but it seems likely to be correlated with greater olfactory and auditory sensory processing and possibly with the development of the highly sensitive mammalian tympanic ear (Kemp, 2005, 2006). These senses would be especially important in the small, nocturnal animals which form the base of the mammalian tree. Interestingly, the cynodont – for example Chiniquodon ¬- brain does not have enlarged cerebral hemispheres, but the cerebellum is enlarged compared to reptiles (or rather amphibians as the modern reptile brain is thought quite derived) (Kemp, 2006), corresponding to the increased necessity for motor control in these upright therapsids. The erect gate is considerably less stable – though allowing much more agility - than the sprawling one, and therefore greater motor control is an absolute necessity. Whether the increased cerebellum of cynodonts was enough to warrant or require endothermy is debatable, but the six-layered neocortex of the mammalian brain has almost certainly been reliant on internal heat production since its inception. Another strand of fossil evidence which has been used to investigate the onset of endothermy has been ecological. As endotherms require much greater amounts of energy, the ratios of predators to prey in the ecosystems of which they form part tend to be smaller. This has thus been attempted to be used as evidence for the timing of occurrence of endothermy. Bakker (1975, 1980) for example has looked at therapsid ecosystems, and concludes based on predator prey ratios that endothermy is indeed present. However, it is debatable whether the fossil evidence is actually good enough to draw such conclusions as it is uncertain to what extent these ecosystems relied on fresh water and whether all herbivorous therapsids were available as prey to carnivores. Towards the end of the therapsids’ ‘reign’ in the Triassic there were increasingly more and more non-therapsid amniote predators, distorting the picture. Thus taking the ecosystem based approach, though intriguing, is of little aid. As can be seen, it is difficult to pin down the time of development of endothermy in synapsids. As far as osteological correlates are concerned a general picture with the latest cynodonts almost certainly being endothermic, the pelycosaurs ectothermic, and the therapsids in between gradually acquiring the features needed to maintain an active metabolism appears possible. Though there is uncertainty earlier in the therpasid and cynodont lineages, early mammals were certainly endotherms, with their possessing all the characters which mark out modern mammals as such, including almost certainly hair insulation, and their immediate cynodont ancestors such and tritheledontids and tritylodontids were likely very similar. The forces actually driving the evolution of homeothermy are somewhat uncertain - not helped by the difficulty in dating onset of homeothermy itself – and there are many theories which attempt to explain them. As explained, modern mammals are generally endothermic homeotherms, as are birds, with their maintaining their internal temperature at a level higher than the ambient temperature. This elevated body temperature is maintained by means of a higher basal metabolic rate (BMR), which is some 6 to 10 times higher than in ectotherms. Elevated temperature and BMR are however of no adaptive use alone, and it is rather the two functions which these allow to be achieved – a constant body temperature and higher aerobic scope – which are the important aspects in terms of selective pressure. Thus theories to explain the origins of mammalian endothermy have nearly all taken one of two basic approaches – aerobic capacity first, or thermoregulation first – with the inherent assumption that one must have preceded the other. As mentioned, a constant body temperature allows a greater degree of complexity to be developed. One example of this complexity is the central nervous system, which is much more developed in mammals than in their reptilian ancestors or in pelycosaurs. This has many advantages including greater neuromuscular control, allowing animals such as tritheledonts and early mammals to be extremely agile, and also to develop very fine control of jaw musculature thereby improving feeding efficiency. Increased sensory perception – especially useful in a nocturnal niche –would also have been allowed by the yet further increases in neocortex size seen in mammals as opposed to cynodonts (see below). Higher aerobic scope also has numerous advantages as it simply allows the animal to be more active for longer periods, thereby increasing its ability to find food and mates, escape from predators, defend its territory, and provision for its young. It does not allow the animal to actually run faster, but it means that high speeds can be maintained for longer periods. There are two key hypotheses as to the origin of endothermy based on a ‘thermoregulation first’ approach. One of these is the ecological, nocturnalization hypothesis. In order to maintain a constant body temperature via metabolism it is necessary to create a temperature gradient between the body and the environment (with the former warmer than the latter). This is easier at night because the external environment is cooler, and thus so too can be the body temperature. Thus in this hypothesis the maintenance of a constant body temperature evolved in response to the nocturnal niche, where external heat is less available, but where maintaining such a gradient is easier. This is supported by the fact that a lower body temperature is seen in many nocturnal mammals such as tenrecs and the desert hedgehog (Crompton, 1978). These animals maintain ‘reptilian’ energetics in that their aerobic metabolic scope is not greatly higher than in a reptile of the same size, but do maintain a constant body temperature via metabolic means –Fig.X. Thus it has been suggested that this could be the hypothetical ancestral state for mammals, providing a bridge to metabolic homeothermy without immediately incurring high metabolic costs. They would thus have later reinvaded the diurnal niche, correspondingly raising their body temperatures to a higher level to allow maintenance of a gradient with the environment and thereby a constant temperature, which would have become essential to their advanced central nervous system. However, such animals as the tenrec appear to have ‘re-adapted’ for a nocturnal lifestyle, reducing their BMR and thereby aerobic capacity in order to save energy. Other nocturnal mammals from bushbabies to European hedgehogs maintain normal mammalian energetics, and diurnal mammals in cold climates do not maintain lower body temperatures. Another ‘thermoregulation first’ theory has been the miniaturization, or physiological theory, first proposed by McNab (1978). He observed small size in early pelycosaurs, giving way to large forms, but with the first mammals having been once more very small. He proposed that inertial homeothermy – as is thought to have existed in the large dinosaurs – was gained to an extent in the large forms, and that this allowed increased activity rates and aerobic metabolism which could subsequently not be lost without great selective cost. The development of endothermy thus came about in order to allow homeothermy to be maintained but at small sizes, where otherwise the large surface area to volume ratio would have prevented homeothermy and thereby all the selective advantages it gives such as increased activity rate and brain size. Endothermy would have to have been acquired first in reasonably large animals, and then miniaturization would then have been free to occur. Hair may have developed along with miniaturization, as the increased surface area: volume ratio would have otherwise led to very great energy losses. However, the trend in sizes pointed out is questioned by Kemp (2005), with even some of the most advanced cynodonts – which show most osteological correlates of endothermy - such as tritylodontids having members with skulls up to 25cm, and earlier therapsid groups such as dicynodonts having very small members despite being likely ectothermic. Though the first mammals were indeed small, it appears at least not to be possible to say that size was the key driver of endothermy. The key hypothesis taking the other main functional aspect of endothermy as primary is the aerobic capacity hypothesis, proposed by Ruben (1979) and augmented by Bennett (1991) and Hillenius (1994). This idea is that it was on higher aerobic scope which the primary selective pressure acted, and that this developed only to be followed by an increased BMR and ultimately thermoregulation. As mentioned, higher aerobic capacity imparts numerous advantages, though all basically involve allowing the animal to move at high speeds (i.e. be highly active) for longer periods, and these are proposed as strong enough advantages to counterbalance the higher metabolic demands. This theory is supported by the development of characters such as an erect gate, secondary palate, diaphragm, and maxillo-turbinals in relatively large therapsids, which in theory should have been large enough not to need endothermy as they could rely in inertial homeothermy, whereby a roughly constant temperature can be maintained by virtue of a relatively low surface area: volume ratio. Thus aerobic metabolism may predate endothermy, but may have caused the increase in BMR as aerobic scope and BMR are correlated. Kemp (2005) however cites Hayes and Garland (1995), stating that increased BMR is in fact a separate consideration from aerobic metabolism, and is caused by more rather than larger mitochondria in viscera rather than muscles. The fossil evidence is also not clear, as maxillo-turbinals for example may not have been well developed. Thus aerobic scope as the driver for endothermy itself can probably be discounted Another approach - the parental provision hypothesis - in at least one form has combined both temperature regulation and aerobic capacity under the umbrella of another factor. This idea is centred on the observation that both mammals and birds display significantly greater levels of parental care than most ectotherms, and that thus there may be a correlation between parental care and endothermy. It has been suggested that endothermy arose as a means of allowing two key aspects of parental care – maintenance of temperature of young and eggs, and also supply of food to young, with Farmer (2000) combining both. A higher and constant body temperature allows the former, preventing young animals dying from temperature changes which could be a problem as their surface area to volume ratios will be very large. A higher aerobic rate allows the latter, meaning that adults can catch and supply more food to their young per unit time. It has also been mooted that the stimulus for increased BMR may have been necessary size increase in the viscera to allow food uptake and processing levels to increase to allow lactation. If this theory were true, then the key selective pressure would be on parental care, rather than endothermic homeothermy itself. However, quite how the other advantages accrued by increased aerobic activity and thermoregulation could possibly be separated from this factor is unclear, and as soft tissues are not preserved the timing of emergence of lactation for example can never be known for certain. Thus the theory remains untestable. Following a lack of resolution with regards such theories, Kemp (2005, 2006) has proposed that a holistic approach is more appropriate. Rather than insisting that either aerobic capacity or thermoregulation or parental care was the key driving factor, one may look at the picture as a whole. The original step along the path to endothermy may have been a mutation increasing the number or size of mitochondria viscera or muscle. Were such an increase in either, greater amounts of food and oxygen would have been needed to support these higher metabolic or aerobic rates, and therefore any adaptation for better food processing or for greater breathing efficiency would be favoured. This in turn would allow yet further increase in metabolic rate, accruing yet more advantages but requiring yet better food finding and oxygen uptake ability. This kind of pattern could be argued to be seen when the stem mammalian fossil record is analysed. Gradually features such as more upright gate, a semi- then fully closed palate, and possibly nasal turbinates accrue, indicating progressively greater metabolic needs and an ever more active lifestyle. Eventually one ends up with a highly developed endotherm in the form of a mammal, with the final ‘touches’ to the mammalian form such as increased brain size facilitated by what had gone before, and perhaps driven by necessities of the nocturnal niche. Kemp points out that it is important to look at the synapsid fossil record, and indeed all fossil forms, as integrated entities, with each adaptation having to keep pace with every other. Developing, for example, a high aerobic scope in the absence of thermoregulation to allow better neuromuscular control and thereby the higher feeding efficiency to fuel higher aerobic metabolism is simply unfeasible – the development of any one trait will be constrained by every other. There is, however, no doubt that full homeothermy developed first in small, noctural carnivores. Why this should be, if homeothermy does not in fact especially advantage them and ideas such as the nocturnalization and miniaturization theories alone are not satisfactory, is unclear. However Kemp (2005) proposes that one reason may be that such animals simply have a higher evolutionary rate than other forms of synapsids. Thus these ‘final’ features of homeothermy simply appeared in this group because the chances of this happening at any point were higher as the evolutionary rate was higher. Interestingly a higher evolutionary rate in such animals appears to correspond to the pattern seen where following extinctions (or perhaps causing them) there are re-radiations of therapsids from small carnivorous forms. Never does a large herbivorous group give rise to a new radiation for example. The reasons for this pattern may be that demographic factors such as relatively small population size, rapid generation time, and low dispersal rate correlated with a metabolically active, small form predispose them to allopatric speciation (Kemp, 2005). Interestingly, this pattern may also be seen in dinosaurs, with small, often carnivorous forms forming the base of several radiations, and size reduction very rare (Sereno, 1997). When exactly and why homeothermy arose in synapsids is thus still uncertain. It is clear that it existed in the earliest mammals, and possibly also in their closest cynodont relatives. When endothermy would have arisen is deeply intertwined with its function. An integrationist approach may well be the best one to take, looking at all the features and aspects of endothermy in conjunction in order to see how it may have arisen genetically, making certain that each intermediate stage would have been a properly integrated organism. However, departing from Kemp (2005), separating basal metabolic rate from aerobic metabolism also seems unnecessary. The two are both linked to mitochondrial activity, and though increased number may predominate in the viscera and size in muscles, a step in either direction would likely affect those in both tissues to some extent. The lack of correlation in some species is much more the exception than the rule. Though the miniaturization hypothesis per McNab certainly has problems, the size of the first mammals must also not be ignored, especially as it parallels the fact that birds developed from some of the smallest of dinosaurs. Though there were small relatively early therapsids which probably lacked endothermy, these may well not have been as active as mammals. Some degree of endothermy may have developed in even large cynodonts. It may not have been needed to maintain high activity levels thanks to inertial homeothermy, but will likely have occurred along with transition to a generally more active lifestyle. This may then have been ‘taken to extremes’ by mammals in order to move to a small-bodied, nocturnal niche. This ‘extreme endothermy’, along with the necessity for improved auditory and olfactory senses in the nocturnal environment would have allowed/driven enlargement of the brain and full development of a highly sensitive ear. In turn it appears indeed likely that endothermy – and an inability to loose it - limited the size of the largest mammals when they returned to daylight even once dinosaurian competition has ended. Such limitation may well also occur in birds, explaining why they too have not have regained the proportions seen among the extinct dinosaurs. References (only those not listed by Kemp given) Pörtner HO (2004) Climate Variability and the Energetic Pathways of Evolution: The Origin of Endothermy in Mammals and Birds. Physiological and Biochemical Zoology 77(6):959–981 Sereno, P. C. 1997. The origin and evolution of dinosaurs. Annual Reviews of Earth and Planetary Sciences 25 : 435-489.